7.2 Resting State fMRI

Question 1

When you are sleeping is your brain at rest?

Answer 1

no! because the brain is never truly at rest - it is never doing nothing

Question 2

What does resting state fMRI measure?
What signal does it measure?

Answer 2

measures the spontaneous brain activity of a person at rest (not doing anything)
measures BOLD signal

Question 3

Why what are the benefits of using resting state fMRI (rs fMRI) in research?

Answer 3

-focusses on connectivity networks of rs
-can be used an as OBJECTIVE clinical biomarker
-can be done in any population
-quick to set up
- very forgiving with signal acquisition -> you can have had different parameters but still get similar results
-networks of rs are very replicable: similar results seen across studies

Question 4

Why is rs fMRI important?

Answer 4

-gives inherent understanding of brain functional organisation

Question 5

What is the Default Mode Network (DMN)?

Answer 5

most widely studied rs network because it is consistently active during passive resting state

Question 6

When is the DMN active?

Answer 6

passive resting state: When an individual does not focus on incoming stimuli and does not perform any attention-demanding task, and the brain is in a rest wakeful state, the default mode network (DMN) is activated

Question 7

What rs fMRI in term of the haemodynamic response?

Answer 7

haemodynamic response is not triggered by tasks set but by spontaneous neuronal activity

Question 8

Are the activation maps of rs networks and task based networks similar or different in their activation patterns?

Answer 8

similar in activation patterns

Question 9

What are some examples of noise in resting state fMRI?

Answer 9

scanner instabilities, physiological noise (breathing, cardiac), subject motion

Question 10

How did researchers prove that rs fMRI was due to neuronal activity?

Answer 10

Laufs et al. 2003: measured rs fMRI with EEG and found regional correlations between fluctuations in the rs fMRI signal and fluctuations in the power of EEG

Question 11

What is functional connectivity?

Answer 11

fc = a temporal correlation between regional fluctuations in cerebral blood flow or BOLD signal

Question 12

What are the typical parameters of a rs fMRI experiment?

Answer 12

T2*-weighted scans
TE = 35ms
TR = 2600ms
pixel dimensions = 1.8x1.8 with slice thickness of 4mm

Question 13

Which software is used to process rs fMRI data?

Answer 13

SPM (CONN Toolbox)
FSL

Question 14

What is the typical pipeline for fMRI data?

Answer 14

get fMRI images -> preprocessing -> denoising -> postprocessing

Question 15

What are the two main types of analysis methods used in rs fMRI?

Answer 15

voxel-based methods: eg seed-based correlation analysis (SBC) and Independent Component Analysis (ICA)
node-based methods: eg ROIs analysis

Question 16

What is the advantages of choosing a voxel-based ICA hypothesis for your rs fMRI experiment?

Answer 16

-finds whole rs networks without having predefined seeds (no apriori hypothesis, more exploratory)

Question 17

What sort of approach does using a voxel-based ICA hypothesis for your rs fMRI experiment give?

Answer 17

(multivariate) voxel based
spatial
exploratory
data driven/model free
looks at global brain changes

Question 18

Why is a ICA hypothesis a spatial approach for a rs fMRI experiment?
What is the equation underlying this approach?

Answer 18

the fMRI data measured (X) is expressed as a set of unknown spatial patterns (S) and the associated time course (A) -> to give spatial maps
(variables are time/space)

X = A * S

Question 19

What does SBC calculate?
Do you need an apriori hypothesis?
Is it an exploratory approach?
Is it a voxel based approach?

Answer 19

-the correlation coefficient between two areas or regions
-yes! need to have hypothesis to determine the specific seed to investigate
-no it is a localised approach
-yes voxel based

Question 20

For SBC and ROI analysis, a connectivity matrix is created out of what values?

Answer 20

Z scores

Question 21

Which method, ICA or SBC, would be better for studying depression using rs fMRI?

Answer 21

ICA is better because depression is linked to a large-scale network dysfunction rather than just a single brain region. Thus, ICA can identify whole rs networks without needing a predefined seed

Question 22

How does the DMN relate to depression?
What analysis method was used to discover the above?

Answer 22

-studies have found hyperCONNECTIVITY of the DMN in depression
-ICA

Question 23

What is the limitation of using SBC?

Answer 23

SBC requires predefined seeds -> may miss broader network activity

Question 24

What is the limitations of using ICA?

Answer 24

harder to interpret and also requires group level analysis

Question 25

What is the limitation of using ROIs for your hypothesis?

Answer 25

relies on predefined ROIs, so you may miss unknown connections

Question 26

What is the advanatage of using SBC?

Answer 26

simple method which is interpretable and focusses on a specific region

Question 27

Which method, SBC ICA ROIs, is good for group COMPARISONS?

Answer 27

ROIs is good for group comparisons

Question 28

What are the advantages of ROIs analysis?

Answer 28

more flexible than SBC good for group comparisons

Question 29

Is ICA hypothesis or data driven?

Answer 29

ICA doesnt need apriori hypothesis -> is data driven

Question 30

What research area are the three different rs fMRI analysis methods best used for?

Answer 30

SBC = specific circuits (amygdala-PFC anxiety) ICA = exploration of large scale network dysfunction (DMN depression) ROIs connectivity = comparing fc between conditions or groups

Question 31

What did the study by Ten et al. 2020 discover about the functional connectivity of painful diabetic polyneuropathy (DPN)?

Answer 31

in-vivo MRI analysis found significantly lower S1 or somatosensory cortical volumes in non-responders compared to responders to treatment treatment responders, the insula showed significantly greater positive functional connectivity with the corticolimbic network We also found the insula exhibiting significantly higher positive connectivity to the corticolimbic circuitry in responders to neuropathic pain treatment. these corticolimbic regions detected are involved in sensory integration and play a pivotal role in emotion and cognitive dimensions of pain shows associations between structural brain changes and treatment response

Question 31

What are some clinical applications of rs fMRI?

Answer 31

MRI is a biomarker of pain in Diabetic Polyneuropathy

Question 32

What are the two pathways involved in pain? what are their roles?

Answer 32

primary somatosensory cortex (S1) = for coding pain intensity and depicting the location of the pain insula cortex = processes emotional aspects of pain

Question 33

What models/hypotheses/approaches are used in fMRI and rs fMRI?

Answer 33

rs fMRI: ICA, ROI-to-ROI, SBC never GLM fMRI: majority GLM, rarely the other 3

Question 34

Why does rs fMRI never use GLM approach?

Answer 34

because rs fMRI focusses on functional connectivity ICA, ROIs, SBC are data-driven or correlational methods which can capture the spontaneous neural activity no predefined task -> no design matrix to put into GLM -> GLM doesnt apply